<div>im not sure we have an implicit fisheye in general<br></div><div><br></div><div>only with line-like-toplogies there is somehow an implicit fisheye (as paket loss probability of every hop multiplicates)</div><div><br></div>
<div>the denser the network the less effect does this packet loss have, (at least as long as we have in fact no used/working mpr)</div><div><br></div><div>regards Markus</div><br><div class="gmail_quote">On Wed, Dec 23, 2009 at 3:36 PM, Sven-Ola Tücke <span dir="ltr"><<a href="mailto:sven-ola@gmx.de">sven-ola@gmx.de</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">Hey,<br>
<br>
also consider: there is some implicit fish eye already build in: the packet<br>
loss will sum up when flooding the LQTC's from node to node. For example: with<br>
an ETX value of > 10 to a distant node, only very few LQTCs will make it...<br></blockquote><div><br></div><div>if a nodes best link ist that bad, than its imho nearly hopeless anyways *G</div><div><br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<br>
Rule of thumb: Fish=On -> small LQ-Interval, Fish=Off -> longer LQ-Interval<br></blockquote><div>btw what is an LQ-interval? (i assume TC-interval)</div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
</blockquote></div><div><br></div><div>Markus<br></div>